1. Field of the Invention
The present invention relates to a three-dimensional image processing technique of generating a texture image from an original three-dimensional model by means of computer processing.
2. Description of the Background Art
In three-dimensional image processing, at the time of adding data of a pattern or the like to the surface of a three-dimensional shaped model, a method which is so-called texture mapping is employed. The texture mapping is performed generally by associating each of a number of polygons constructing a three-dimensional shaped model with a texture image as a two-dimensional image. Consequently, at the time of performing the texture mapping, it is desired to preliminarily generate texture images corresponding to the parts of a three-dimensional model.
There are mainly two conventional methods of generating a texture image.
In the first method, a cylindrical plane is set around a three-dimensional model on which a pattern or the like is formed, and the pattern on the surface of the three-dimensional model is projected onto the cylindrical plane, thereby generating a texture image as a two-dimensional image.
In the second method, a micro texture image assigned to each of polygons (hereinafter, referred to as an “unit texture image”) is extracted in a form so as to correspond to each polygon shape, and micro texture images are arranged in a plane, thereby generating a texture image as a two-dimensional image.
The conventional texture image generating methods, however, have the following problems.
First, in the case of the first method, all of the parts of a three-dimensional model cannot always be projected onto the cylindrical plane which is set around the three-dimensional model. Consequently, the first method has such a problem that a texture image cannot be generated for a polygon which is in the shade when seen from the cylindrical plane, and cannot be projected.
In the case of the second method, since the unit texture images corresponding to the polygons, which are arranged in a plane, become a texture image, there is the possibility that an image dropout occurs around a boundary of polygons at the time of adding the unit texture images to the polygons. Consequently, in the case of the second method, it is necessary to prevent the dropout around the boundary of polygons by setting the size of each unit texture image to be slightly larger than the size of a corresponding polygon. However, when unit texture images each having a size larger than the polygon size are held for all of polygons constructing a three-dimensional shaped model, a problem such that the size of a whole texture image becomes large arises.
Further, the second method has such a problem that, since the unit texture images arranged in a plane become a texture image, even if the whole texture image is displayed, it is difficult to grasp the general pattern or the like of the three-dimensional model. Consequently, when the user tries to perform an image correcting process or the like on the texture image by activating image processing software or the like, the user cannot grasp the whole image of the three-dimensional model, so that a problem arises such that it is difficult to perform a general correcting work.